How long does it take to accrue decent amounts of data?

I had to move house a few months ago and with the resulting downtime I was disqualified etc. Long story short, I’ve started up a new node. It’s been running for a couple of months now and my 9.5TB enterprise-grade SAN disk hasn’t even cracked 1TB of storage in use. Is this normal?

I’ve got a few 10’s of TB’s sitting offline in disk shelves that I could connect up, but because the storage is so slow to fill up I’m wondering whether it’ll ever be needed.

Some months ago we saw ingress of up to 2.5TB/month (that’s when I bought a new HDD…) but it quickly dropped after that to <1TB/month (current month might be ~650GB). So currently your 9.5TB will take a long time to fill up, just like my remaining 5TB…
And since you have a new node, it took at least a month for vetting during which it got only 5% data.

That’s a bit frustrating. Last month’s bandwidth throughput has been nothing to write home about - averaging 4Mbps each way.

This is my graph of used space vs time (ignore the gaps :slight_smile: ):

As you can see it’s not that steady.

Here is a different type of graph (shorter time range though) - used space change per hour (it’s negative if data is deleted):

The large negative spikes are most likely data from the old stefan-benten satellite being deleted.

Portion of the graph from the spikes to now:

Network traffic:

I have like 250GB/month ingress, at current rates my new 12TB node will die before its full.

My bandwidth over the last month. This nodes only been running 2.5 months, so don’t have an annual graph worth displaying.

you sure you don’t share a subnet with another node…

Storj really needs to get new huge customers on board asap.


There’s not one on my network. I have a /29 of public address space, and there isn’t another node on that subnet. This node is a replacement for my old node that got disqualified when I moved house, but should that affect a new node?

no a node that goes offline will not be counted after 4 hours…

the subnet that data is distributed to globally is based on ip/24 subnets
i really cannot remember my binary math on these masks, was never really into that much…
basically it’s if there is any other nodes in a subnet with you then you will be sharing with them…

It’s always possible there are other people on the /24 that my subnet is part of, but I wouldn’t have thought it was likely. Perhaps I could create another node on a different subnet and see how that goes.

many will use the default port for storj… so you can scan the subnet in that range for an easy basic overview if there are any on your particular subnet

Done some port scans using nmap. Nothing other than me is open on 28967 in my subnet.

1 Like